Comprehensive cardiac evaluation by perfusion CMR improves risk prediction in chest pain patients compared to myocardial perfusion SPECT
M Kraen, S Akil Engblom, M Permlid, Barbro Kjellstrom, Ellen Ostenfeld, Peter Kellman, Hakan Arheden, Henrik EngblomAbstract
Background
Comprehensive cardiac evaluation of ischemia, infarction and left ventricular ejection fraction (LVEF) can be achieved with perfusion CMR (pCMR) or gated myocardial perfusion SPECT (MPS). Both modalities have shown prognostic value in patients with chronic coronary syndrome (CCS). However, comparisons based on large observational studies are absent.
Purpose
To compare the predictive value of comprehensive cardiac evaluation by pCMR and MPS for risk of major adverse cardiovascular events (MACE).
Methods
A single-center study of patients >18 years referred for pCMR or MPS between 2016 and 2022. The clinical reports were used for determining the presence of ischemia, infarction and the calculated LVEF. Imaging results were considered abnormal if ischemia or infarction was present or if LVEF ≤45%. Patients were followed for up to five years, with outcome data obtained from national health data registries. MACE consisted of cardiovascular death, myocardial infarction or any coronary intervention performed more than 6 months after the index exam.
Results
In the study cohort (n= 4405) the mean age was 64 years, 46 % were female and 22% had prior ischemic heart disease (IHD).Patients in the pCMR group (n=2117) were on average 3 years older. Imaging findings of ischemia, infarction and LVEF ≤ 45% were all more commonly detected by pCMR compared to MPS (39% vs 29%, p<0.001). During a mean follow up of 3.2 years MACE occurred in 147 (6.9%) pCMR patients and 187 (8.2%) MPS patients (Table 1). The Kaplan-Meier curves (Figure 1) illustrate the improved risk discrimination by pCMR. In Cox regression models adjusting for age, sex, comorbidities and cardiac medication, the hazard ratios (HR) for abnormal imaging findings were: HRpCMR = 3.8, 2.5-5.8, p<0.001 and HRMPS = 1.6, 1.2-2.2, p=0.002. An interaction analysis of the pooled data set (n=4405) showed a significant interaction of modality and abnormal findings (p=0.007) proving that risk prediction differed depending on modality.
Conclusions
A comprehensive cardiac evaluation of ischemia, infarction and LVEF by pCMR is superior compared to MPS for predicting future ischemic events. If functional testing is performed in patients with CCS, pCMR should be the method of choiceTable1Figure1